Decision support systems in clinical practice: The case of venous thromboembolism prevention.

BACKGROUND: Today medicine is facing a "knowledge crisis" in that explosively expanding medical knowledge encounters limited abilities to disseminate new practices [1]. Clinical practice guidelines (CPGs) are intended to promote high standards of care in specific areas of medicine by summarizing best clinical practice based on careful reviews of current research. However, doctors are often short of time to study these documents and check their updates, have little motivation for strict adherence to them. A systematic review of 11 studies reporting on 29 recommendations has found that median adherence to all recommendations was 34%, suggesting that potential benefits for patients from health research may be lost [2].Clinical decision support systems (CDSS) can serve as a knowledge translation tool, mediator between clinical guidelines and physicians by providing the right information to the right person at the right time. OBJECTIVE: To evaluate the effectiveness of implementation of international and national CPGs for venous thromboembolism (VTE) prevention with the help of CDSS in a general hospital. METHODS: A multifunctional CDSS based on national and international guidelines on the VTE prevention was developed and implemented in the Medical Center of the Bank of Russia (MC). The system has the following functionalities: 1) it supports the decision on the VTE prevention based on individual risk assessment of thrombosis (scales of Caprini, Rogers and Khorana, Padua Prediction Score, additional risk factors) and bleeding (IMPROVE scale for non-surgical patients, major bleeding scale for surgical patients and major orthopedic surgeries, hemorrhagic complications risk in cancer patients); 2) generates the summary containing the grade of recommendations and the level of evidence, personalized recommendations on regimen and duration of preventive antithrombotic therapy, dose correction according to creatinine clearance; 3) provides an audit form for and statistical analysis of VTE cases; 3) automatically generates a quality register for VTE prevention.CDSS was implemented in June 2014. We analyzed VTE cases identified by triggers (deep vein thrombosis diagnosed by Doppler ultrasound and pulmonary embolism at the chest CT) that occurred in 2014 before and after CDSS implementation, as well as in the first half of 2015. Patients with VTE diagnosed during the first 48 hours of hospitalization or receiving anticoagulants in therapeutic doses were excluded from the analysis. Chi-square test for linear trend and non-parametric methods of descriptive statistics were used for data analysis. RESULTS: CDSS utilization was regulated by a special hospital-wide policy; lectures were organized to educate doctors how to use the system. Although international recommendations require VTE risk assessment for all hospitalized patients (except those receiving anticoagulant in therapeutic doses), the doctors filled forms for only 306 patients during the first 6 months of CDSS functioning (14.1% of discharges with length of stay >48 hours during this period). In the first half of 2015 the coverage of VTE risk assessment with CDSS was 19% (n = 506). Correctness of filling out the forms was 78.4%, in the rest of cases doctors made mistakes in choosing patient's profile or when filling in risk scales.Doctors adhere to given recommendations in 85.4% of cases. Most often (47.5%) pharmacotherapy with low molecular weight heparin (LMWH), preventive doses, was recommended by the system, and in this category the adherence to recommended practice was the lowest (74.6%). Among patients who underwent pharmacoprophylaxis, in 21.1% cases the use of anticoagulants was inconsistent with clinical guidelines or drug package insert (typically inappropriate choice of LMWH prophylactic doses, delaying or reducing the duration of prophylaxis).The rate of hospital-acquired VTE significantly decreased after CDSS implementation and was 11.71, 8.28 and 4.84 per 1,000 hospitalizations in the first and second half of 2014 and in the first half of 2015, respectively (χ2 = 7.325, df = 1, p = 0.0068). The rate of postoperative VTE for the same period amounted to 8.76, 3.39 and 4.17 per 1,000 operations, respectively (χ2 = 7.266, df = 1, p = 0.007), reaching a level of the correspondent AHRQ safety indicator (4.99 per 1,000 operations) [3]. Deviations from clinical guidelines or anticoagulant package inserts were revealed in 74% of VTE cases; and more than 1/3 of deviations affected treatment outcomes. CONCLUSIONS: Coverage of hospitalized patients with documented VTE risk assessment gradually increased after the CDSS implementation, but remained at a low level (19% of eligible patients). Partly it may be attributed to the lack of CDSS integration in electronic health record or computerized physician order entry systems that would facilitate routine documentation of VTE and bleeding risks. However, the introduction of CDSS has allowed reducing significantly the rate of hospital-acquired VTE. This can be explained by drawing doctor's attention to the VTE problem and by training effect of CDSS. After receiving appropriate recommendations doctors adhere to them, on average, in 85.4% of cases, although for LMWH pharmacoprophylaxis this level was lower (74.6%). Development of hospital-acquired VTE in most cases (74%) was accompanied by non-compliance with CPGs recommendations, emphasizing the importance of additional measures for better adherence to evidence-based clinical practices.

Tuberin-dependent membrane localization of polycystin-1: a functional link between polycystic kidney disease and the TSC2 tumor suppressor gene.

The PKD1 gene accounts for 85% of autosomal dominant polycystic kidney disease (ADPKD), the most common human genetic disorder. Rats with a germline inactivation of one allele of the Tsc2 tumor suppressor gene developed early onset severe bilateral polycystic kidney disease, with similarities to the human contiguous gene syndrome caused by germline codeletion of PKD1 and TSC2 genes. Polycystic rat renal cells retained two normal Pkd1 alleles but were null for Tsc2 and exhibited loss of lateral membrane-localized polycystin-1. In tuberin-deficient cells, intracellular trafficking of polycystin-1 was disrupted, resulting in sequestration of polycystin-1 within the Golgi and reexpression of Tsc2 restored correct polycystin-1 membrane localization. These data identify tuberin as a determinant of polycystin-1 functional localization and, potentially, ADPKD severity.

Determination of loss of heterozygosity in frozen and paraffin embedded tumors by denaturating high-performance liquid chromatography (DHPLC).

Spontaneous renal cell carcinoma (RCC) occurs with a high frequency in Eker rats carrying a germline alteration of the tuberous sclerosis-2 (Tsc-2) tumor suppressor gene. To determine the frequency with which the wild-type allele of the Tsc-2 gene is lost in RCC and the ability of DHPLC to detect loss of heterozygosity (LOH) at this gene locus, fresh-frozen and paraffin-embedded formalin-fixed tumors from heterozygous Eker rats (Tsc-2(Ek/+)) were examined for LOH at the Tsc-2 locus. LOH was determined by quantitation of peak areas of PCR products specific for the mutant and wild-type Tsc-2 alleles. For normal DNA isolated from heterozygous animals, the allele ratio (AR) of mutant to wild-type PCR products was empirically determined to be 1.5+/-0.3 (n=30) and LOH was defined as >2 standard deviations away from this mean, i.e. any AR >2.1. Analysis of 15 spontaneous frozen RCC samples showed LOH in 10/15 samples (66%). Carcinogen-induced tumors exhibited an even higher frequency of LOH, with 6/6 paraffin-embedded, formalin-fixed tumors exhibiting LOH. 100% concordance was observed between the results obtained by DHPLC and traditional methodologies. Therefore, LOH appears to occur with a high frequency in both spontaneous and carcinogen-induced RCC in this animal model and DHPLC is a sensitive and high throughput methodology for detecting this type of genetic alteration.

Carcinogenicity of a nephrotoxic metabolite of the "nongenotoxic" carcinogen hydroquinone.

Hydroquinone (HQ) is a potential human carcinogen to which many people are exposed. HQ generally tests negative in standard mutagenicity assays, making it a "nongenotoxic" carcinogen whose mechanism of action remains unknown. HQ is metabolized to 2,3,5-tris(glutathion-S-yl)HQ (TGHQ), a potent toxic and redox active compound. To determine if TGHQ is a carcinogen in the kidney, TGHQ was administered to Eker rats (2 months of age) for 4 or 10 months. Eker rats carry a germline mutation in the tuberous sclerosis 2 (Tsc-2) tumor suppressor gene, which makes them highly susceptible to the development of renal tumors. As early as 4 months after the initiation of treatment (2.5 micromol/kg, i.p.), TGHQ-treated rats developed numerous toxic tubular dysplasias of a form rarely present in vehicle-treated rats. These preneoplastic lesions are believed to represent early transformation within tubules undergoing regeneration after injury by TGHQ, and adenomas subsequently arose within these lesions. After treatment for 10 months (2.5 micromol/kg for 4 months followed by 3.5 micromol/kg for 6 months), there were 6-, 7-, and 10-fold more basophilic dysplasias, adenomas, and renal cell carcinomas, respectively, in TGHQ-treated animals than in controls. Most of these lesions were in the region of TGHQ-induced acute renal injury, the outer stripe of the outer medulla. Loss of heterozygosity (LOH) at the Tsc-2 locus was demonstrated in both the toxic tubular dysplasias and tumors from rats treated with TGHQ for 10 months, consistent with TGHQ-induced loss of tumor suppressor function of the Tsc-2 gene. Thus, although HQ is generally considered a nongenotoxic carcinogen, our data suggest that HQ nephrocarcinogenesis is probably mediated by the formation of the quantitatively minor yet potent nephrotoxic metabolite TGHQ, which induces sustained regenerative hyperplasia, loss of tumor suppressor gene function, and the subsequent formation of renal adenomas and carcinomas. In addition, our data demonstrate that assumptions regarding mechanisms of action of nongenotoxic carcinogens should be considered carefully in the absence of data on the profiles of metabolites generated by these compounds in specific target organs for tumor induction.

Tissue-specific expression and splicing of the rat polycystic kidney disease 1 gene.

Autosomal dominant polycystic kidney disease (ADPKD) is the most common genetic potentially lethal human disorder and the polycystic kidney disease 1 (Pkd1) gene is accounted for 85-90% of these cases. We have obtained rat Pkdl cDNA sequence and characterized splicing of Pkdl RNA transcripts in normal rat tissues. Our sequence data revealed a high conservation of the Pkdl gene between rat and other species and mapped rat Pkdl to chromosome 10 in "tail-to-tail" orientation to the tuberous sclerosis 2 (Tsc2) gene. Pkdl was found ubiquitously expressed in the normal rat tissues and the brain had a complex pattern of exon 12 splicing. A novel splicing variant lacking entire exon 31, which occurs in rat and mouse but not in humans, was also identified. As the rat appears to be a valuable model for investigating polycystic kidney disease, the characterization of the rat Pkdl gene will help facilitate future studies to elucidate the molecular mechanisms of cystogenesis in this animal model.

Application of high-performance liquid chromatography-based analysis of DNA fragments to molecular carcinogenesis.

Denaturing high-performance liquid chromatography (DHPLC)-based DNA fragment analysis is a high-throughput technology that can be used to obtain information on both genetic alterations and gene expression. By using different approaches based on polymerase chain reaction, this technique can be used to determine loss or gain of an allele, to quantitate the amount of RNA expressed, and to detect a single nucleotide change. Applications of DHPLC to molecular carcinogenesis include genotyping of transgenic animals; determination of allelic imbalances, including loss of heterozygosity in tumors; measurement of changes in gene expression; and detection of DNA polymorphisms and point mutations. In our laboratories DHPLC has been validated and used to genotype an Eker rat colony, to study the genetic profile of renal cell carcinomas, to quantitate expression of the keratinocyte lipid-binding protein gene in 8-lipoxygenase transgenic mice, and to detect polymorphisms and a point mutation in the tuberous sclerosis 2 tumor suppressor gene in t-haplotype mice.

Requirement for the von Hippel-Lindau tumor suppressor gene for functional epidermal growth factor receptor blockade by monoclonal antibody C225 in renal cell carcinoma.

Renal cell carcinoma (RCC) is a cytologically and histologically diverse disease in which a spectrum of distinct molecular alterations occurs, including the inactivation of the von Hippel-Lindau (VHL) tumor suppressor gene, which is specific for the clear cell variant of RCC. The prognosis for RCC is poor, and, to date, no effective systemic treatment is available for this cancer. In the present study, we assessed the extent to which the transforming growth factor alpha-epidermal growth factor receptor (EGFR) autocrine loop could be used as a potential therapeutic target for RCC. Northern blot analysis of transforming growth factor alpha and EGFR revealed variable but consistent expression of these transcripts in cell lines derived from both clear cell and non-clear cell RCC variants, indicating the potential for this autocrine loop in both tumor types. The therapeutic utility of interruption of this feedback loop was determined by examining growth inhibition after the exposure of these cell lines to a humanized anti-EGFR monoclonal antibody, C225. In vitro treatment of clear cell RCC-derived cell lines lacking VHL resulted in only a modest decrease in growth rate. In contrast, non-clear cell RCC-derived cell lines that retained VHL responded significantly to C225 treatment. Transfection of VHL into VHL-negative RCC cell lines restored responsiveness to C225, indicating that this tumor suppressor gene is required for effective EGFR blockade. Growth inhibition by C225 in VHL-positive cells was linked to a requirement for VHL to up-regulate p27 in response to C225. These data provide compelling evidence that treatment modalities for RCC are likely to be strongly influenced by the molecular etiology of this phenotypically diverse cancer.

Genetic variants of the tuberous sclerosis 2 tumour suppressor gene in mouse t haplotypes.

The murine t complex on chromosome 17 contains a number of homozygous lethal and semi-lethal mutations that disrupt development of the mouse embryo. We recently characterized an embryonic lethality in the rat that results from a germ-line mutation in the tuberous sclerosis 2 (Tsc-2) tumour suppressor gene (the Eker mutation). Remarkably, mouse embryos homozygous for tw8 mutation display cranial defects reminiscent of those observed in rat embryos homozygous for the Eker mutation. To determine whether the Tsc-2 gene, which is in the t complex, is mutated in tw8 or other t haplotypes, we characterized this gene in a series of t haplotype mice. Four Tsc-2 polymorphisms were identified: three in the coding region and one intronic that appeared to be common to all t haplotypes analysed. No evidence was found to argue that the Tsc-2 gene is altered in tw8 haplotype mice. However, in the tw5 haplotype we found a G to T mutation in Tsc-2 that was present only in this t haplotype. In contrast to other polymorphisms within the Tsc-2 coding region which did not result in amino acid changes in Tsc-2 gene product tuberin, this mutation substituted a phenylalanine for a conserved cysteine in tw5 tuberin. Within the t complex, the Tsc-2 gene and the putative tw5 locus appeared to map to different positions, complicating identification of Tsc-2 as a candidate for the tw5 locus and suggesting that the G to T mutation in the Tsc-2 gene may have arisen independently of the tw5 functional mutation.

Mesotheliomas induced in rats by the fibrous mineral erionite are independent from p53 alterations.

The development of human malignant mesothelioma (MM) is strongly associated with occupational or environmental exposure to certain natural mineral fibers, although the genetic mechanisms underlying this malignancy remain unclear. Although the p53 gene is frequently mutated in various tumors, human asbestos-associated MMs appear to develop independently from p53 alterations. The high mesotheliomagenic potency of natural fibrous mineral erionite is well established in humans and rodents, but no data regarding genetic alterations in erionite-associated tumors are currently available. Previous speculations that the oncogenic mechanisms underlying asbestos and erionite carcinogenesis may differ led us to examine whether the p53 gene is targeted in erionite carcinogenesis. Fifteen erionite-induced rat MMs as well as six cell lines derived from asbestos-induced and spontaneous rat MM were analyzed for p53 mutations by direct DNA sequencing and immunohistochemical analysis. Both approaches did not reveal p53 alterations in rat MM samples used in the study indicating that, similar to asbestos carcinogenesis, erionite carcinogenesis does not target the p53 tumor suppressor gene.

Loss of function of the tuberous sclerosis 2 tumor suppressor gene results in embryonic lethality characterized by disrupted neuroepithelial growth and development.

Germline defects in the tuberous sclerosis 2 (TSC2) tumor suppressor gene predispose humans and rats to benign and malignant lesions in a variety of tissues. The brain is among the most profoundly affected organs in tuberous sclerosis (TSC) patients and is the site of development of the cortical tubers for which the hereditary syndrome is named. A spontaneous germline inactivation of the Tsc2 locus has been described in an animal model, the Eker rat. We report that the homozygous state of this mutation (Tsc2(Ek/Ek)) was lethal in mid-gestation (the equivalent of mouse E9.5-E13.5), when Tsc2 mRNA was highly expressed in embryonic neuroepithelium. During this period homozygous mutant Eker embryos lacking functional Tsc2 gene product, tuberin, displayed dysraphia and papillary overgrowth of the neuroepithelium, indicating that loss of tuberin disrupted the normal development of this tissue. Interestingly, there was significant intraspecies variability in the penetrance of cranial abnormalities in mutant embryos: the Long-Evans strain Tsc2(Ek/Ek) embryos displayed these defects whereas the Fisher 344 homozygous mutant embryos had normal-appearing neuroepithelium. Taken together, our data indicate that the Tsc2 gene participates in normal brain development and suggest the inactivation of this gene may have similar functional consequences in both mature and embryonic brain.

Identification of a tumor-specific methylation site in the Wilms tumor suppressor gene.

Malignant mesothelioma is one of the very few extrarenal neoplasms in which the Wilms tumor suppressor gene (wt1) is expressed. We examined wt1 for alterations in rat mesotheliomas, a well characterized animal model for the human disease. Southern analysis revealed a 3.5 kb EcoRI wt1 fragment readily detectable in majority of mesothelioma cell lines and primary mesotheliomas but not in normal rat tissues. Cloning and sequencing of this fragment revealed that the presence of this EcoRI fragment resulted from an inability of this enzyme to cut at a EcoRI site in intron 1 of wt1. This site contains potential motifs for cytosine methylation and treatment of mesothelioma cells with 5-azadeoxycytosine restored the normal EcoRI digestion pattern of wt1 in these cells indicating that cleavage was inhibited by methylation at this site. Southern analysis using HpaII/MspI digestion revealed no differences in methylation between mesothelioma cell lines and normal mesothelium at other CpG sites in wt1 5' region. Renal cell carcinoma lines which did not express wt1 were also methylated at this EcoRI site. Our identification of a site frequently methylated in malignant cells, independent of gene expression, provides a new model system to study determinants of site-specific methylation in tumors.

Characterization of the rat neurofibromatosis 2 gene and its involvement in asbestos-induced mesothelioma.

The neurofibromatosis 2 (NF2) tumor suppressor gene was recently implicated in the genesis of human mesothelioma. To investigate the role of this tumor suppressor gene in rat asbestos-induced mesothelioma, a commonly used model for the human disease, we characterized the rat homologue of NF2 and examined rat chrysotile-induced primary mesotheliomas and cell lines derived from chrysotile- and crocidolite-induced mesotheliomas for alterations in this gene. The coding sequence obtained for the rat NF2 gene had 90% nucleotide homology with the human NF2 gene. The rat NF2 gene was ubiquitously expressed as a 4.4-kb transcript in normal rat tissues as well as in rat mesothelioma cell lines. Reverse transcription-polymerase chain reaction analysis to examine splicing of NF2 exons in mesothelioma cells indicated that the exon splicing pattern was similar in normal and neoplastic cells. To determine if mutations had occurred in the NF2 coding region in rat mesotheliomas, single-strand conformation polymorphism analysis and direct sequencing were used to screen 10 primary tumors and six tumor cell lines. No DNA sequence alterations were observed in any of the rat mesothelioma samples examined. These findings contrast with data reported previously for human mesotheliomas, in which the NF2 gene was found to be mutated in 40% of cases. Taken together, these data suggest that the role of NF2 in the development of rodent asbestos-induced mesothelioma may differ significantly from the role in the human disease.